Diagnostic accuracy of VIA and HPV detection as primary and sequential screening tests in a cervical cancer screening demonstration project in India.
Many limited-resourced countries have either introduced cervical cancer screening programs or are contemplating to do so using visual inspection after acetic acid application (VIA) or human papillomavirus (HPV) detection tests. Both tests have high false-positivity and a suitable triaging strategy is required. Colposcopy triaging is not practicable in most resource-limited settings due to several reasons. We evaluated a portable, battery-operated, magnifying device (GynocularTM) to triage screen positive women in community setting in India. Women positive on VIA or oncogenic HPV test were examined with Gynocular by clinicians in primary health clinics. Findings were documented using the International Federation for Cervical Pathology and Colposcopy (IFCPC) terminology. Swede score was also calculated. Biopsy was performed irrespective of Gynocular findings. The accuracy of Gynocular to detect high-grade lesions or cancer (HSIL+) was estimated. The suitability of Gynocular to correctly triage screen positive cases for immediate ablative treatment was also evaluated by creating simulated scenarios. Sensitivity and specificity of Gynocular were 96.4 and 47.1 %, respectively, to detect HSIL + at the threshold of IFCPC grade 1 findings. Increasing threshold to grade 2 changed sensitivity and specificity to 92.9 and 94.1 %, respectively. Optimum combination of sensitivity and specificity as determined by the receiver operating curve analysis was at the cut-off Swede score of 5. Triaging of VIA/HPV positive women to treatment using grade 2 criteria would have resulted in modest overtreatment and missing of very few high-grade lesions. Gynocular can be used as an effective triaging device for VIA/HPV positive women.